A portfolio-wide operating model change across four value streams

Healthcare distribution and its technology arms are being reshaped from the top down. Capital is moving toward oncology and specialty services, low-growth businesses are being separated rather than repaired, and segment reporting has been redrawn around the parts that are growing. A technology business sitting inside one of those parents inherits every one of those decisions and controls none of them.

The market it sells into moved at the same time. Third party sector research, not a measurement of this engagement: KLAS Research’s 2026 acute care EHR market share report, covering calendar 2025, found the number of hospitals affected by new EHR purchase decisions fell roughly 40 percent against 2024 and nearly 50 percent against 2023, with capital redirected toward AI and operational efficiency. Platform selection is mostly finished. What health systems buy now is more value out of what they already run, which changes what a healthcare technology portfolio is supposed to produce.

Their customers have very little room to absorb a wrong guess. Also third party research about the sector: Kaufman Hall’s national analysis, drawn from more than 1,300 hospitals, put the calendar 2025 median hospital operating margin at 1.3 percent including allocations, and the American Hospital Association’s 2025 Cost of Caring report gives hospital expense growth of 5.1 percent in 2024 against 2.9 percent general inflation.

So the portfolio question gets sharper at the same moment the portfolio decision gets slower. Funding arrives on the parent’s fiscal calendar. Procurement runs centrally. The people who understand the product best are no longer the people who decide what it gets funded to do.

A healthcare technology business inside a much larger parent had the problem that shows up when a fast company is acquired by a slow one. Four value streams were each doing sensible things, and no single view existed of what the whole portfolio was committed to. Decisions that used to take a conversation now took a cycle.

None of the four was failing, which is what made it awkward to name. Each ran its own planning rhythm, described its work in its own vocabulary, and held its own read on what mattered next. Every one of those was defensible on its own. Together they were unreadable. Leadership could not see in one place what the portfolio had already said yes to, so each new commitment was made without anyone knowing what it displaced.

The buyer was a director who had to escalate to a senior director to get the work approved, which tells you most of what you need to know about the funding environment they were operating in. The constraint was not craft. The teams knew how to build. The constraint was that the authority to fund had moved away from the people holding the product knowledge, and nothing in the operating model had been rebuilt to close that distance.

Leadership first, in two stages, in that order deliberately. A private SAFe® Lean Portfolio Management programme for the leaders who owned funding decisions. Then Agile Product Management for the far larger product community who had to turn those decisions into roadmaps.

The order follows the constraint. If the bottleneck sits at the funding decision, the first population to change has to be the one that makes it. Running those the other way round is the common and expensive mistake. Teaching product managers to plan against a portfolio that has not agreed how it funds things produces frustration rather than roadmaps. It also burns the credibility of the method with the exact population you need on side, and that credibility does not come back on a second attempt.

Three customised leadership sessions in between. Not repeats of the class. Working sessions built on the specific decisions the portfolio was actually stuck on, using live disagreements as the material. This is where a framework either earns its place or gets quietly abandoned after the certificates arrive. Standard content teaches people a vocabulary. A working session on a real, contested funding decision teaches them whether the vocabulary changes the answer. If it does not, they are right to drop it.

Then we embedded. Coaches inside the four value streams, badged, on the client’s premises, in the teams’ own ceremonies rather than in a weekly advisory slot. Sustained across all four streams at once, running alongside tooling and value stream enablement.

Embedding costs more than advising and works better for a reason that is easy to state and hard to replicate. A coach sitting in a review hears the argument that never makes it into the written summary. They become someone the team asks rather than someone the team reports to, and those are different jobs with different information available to them. An advisor gets the version of events the team has already tidied up. By the time a problem is legible enough to appear in a status report, the decision that caused it is several weeks old.

Four value streams concurrently, not sequentially. Sequential coaching across a portfolio produces four different local dialects of the same method and a leadership team receiving four incompatible reports. It also guarantees that stream one has drifted before stream four has started. Running them together is harder to staff and is the only way the portfolio view at the top is built from comparable parts. That comparability is the whole product. A portfolio view assembled from four differently defined inputs is worse than no view, because it looks authoritative.

We structured it as a fixed commitment against a defined scope rather than an open-ended hourly arrangement.

For a client whose parent runs procurement centrally and whose financial year does not align with the calendar, an hourly arrangement would have meant a recurring negotiation about time, in an organisation where every one of those goes through a governance step. A fixed structure meant their finance function could plan with certainty, and our coaches could spend their attention on the value streams instead of on administration.

The tradeoff is real and it runs against us. A fixed structure means we absorb what scope discovery finds, and embedded coaching always finds more than the scoping conversation predicted. We carried the delivery risk rather than the client. That is the correct place for it to sit when we are the ones who scoped the work, and it is the reason the programme survived a financial year boundary that would have ended an hourly arrangement.

Start where the decision is made, not where the work is done. When a business sits inside a larger parent, the slow part is almost never the build. Any organisation that finds itself training teams harder while the funding queue grows is solving the wrong end.

Run the streams together. The reason to accept the harder staffing problem is that a portfolio view is only worth having if its parts were defined the same way, and definitions drift the moment you sequence.

Put the coaches where the argument happens. Advisory access gives you the summary. Presence gives you the disagreement, and the disagreement is the useful part.

Structure the commitment so it survives the client’s own calendar. In an environment where every purchase clears a governance step, an arrangement that needs renegotiating on a cadence will eventually meet a boundary it does not clear. That is a delivery design question before it is a commercial one.

And do not wait for the restructuring to end. In this sector, a business that only changes how it decides during calm periods will not change at all.

The client initiated the extension. It started inside their own governance process rather than with a proposal we pushed. In an environment where every purchase has to survive a governance cycle, a buyer starting that process unprompted is the strongest available signal, and a more honest one than a satisfaction score.

The work is live and coaching continues across all four value streams.